Content uploaded by Michael Wittmann
Author content
All content in this area was uploaded by Michael Wittmann on Oct 26, 2018
Content may be subject to copyright.
A preview of the PDF is not available
Agent-based Modeling and Simulation of Electric Taxi Fleets
Abstract and Figures
Electrification and on-demand services are one of the main driving forces within the current automotive sector. This paper presents an approach to modeling and simulation of on-demand applications on the example of an electric taxi fleet. With regard to the high daily mileage and just the same idle times, the characteristic mobility behavior of taxis offers ideal conditions for electrification. To support decision-making during strategic and operational planning, this paper suggests a stochastic model based on an agent-based simulation approach. The simulation engine consists of an event-driven architecture. Customer demand requests, a customizable fleet configuration and infrastructure settings form the main input interface. The simulation output describes use of the infrastructure and the spatial and temporal behavior of each agent. We verify our basic model design first with a combustion engine powered taxi fleet. An additional scenario with electric vehicles provides insights into feasible electrification strategies for the taxi system in Munich. The strength of the proposed model is its distributed, behavior-driven architecture. This is especially useful for on-demand fleets, as these mobility systems are characterized by a mixture of centralized and decentralized knowledge bases. The whole system behavior results from dynamic decision making. Our approach can be used to evaluate various mobility-as-a-service concepts.
Figures - uploaded by Michael Wittmann
Author content
All figure content in this area was uploaded by Michael Wittmann
Content may be subject to copyright.
Supplementary resource (1)
... Agent-based electromobility simulations can be implemented within a wide range of simulation frameworks and using various modeling techniques. In the past, such simulations were realized as Monte-Carlo [12], discrete event-based [17,26,27], and, most prominently, microscopic transport simulations [13,14,18,19,[28][29][30][31]. While Monte-Carlo approaches and discrete event-based simulations are commonly realized in the form of custom implementations, authors typically adapt existing simulation frameworks for microscopic transport simulations, due to their complexity. ...
... Hidalgo and Trippe [17], for instance, assume charging to take place whenever an agent parks at a destination with an SOC of below 30% or when their SOC falls below 60% and the parking duration is longer than 2 h. Similarly, Jäger et al. [26,27], Bischoff and Maciejeweski [28], and Zhang et al. [31] stipulate fixed charging thresholds. Threshold-based charging behavior models users who avoid range anxiety and possible car break-downs long before they appear. ...
... In the vast majority of studies with explicit location choice, charging is modeled to take place at the nearest charger as soon as the respective charging behavior calls for recharging [12][13][14]19,28,30,31]. In simulations in which trips cannot be interrupted or in which chargers are available at only a few, dedicated facilities, agents charge at their trip destinations [15,17] or at the nearest charging station after completing their trip [26,27]. The aforementioned studies only consider proximity in the context of choosing a specific charger, and any other potential influences are left out. ...
In order to electrify the transport sector, scores of charging stations are needed to incentivize people to buy electric vehicles. In urban areas with a high charging demand and little space, decision-makers are in need of planning tools that enable them to efficiently allocate financial and organizational resources to the promotion of electromobility. As with many other city planning tasks, simulations foster successful decision-making. This article presents a novel agent-based simulation framework for urban electromobility aimed at the analysis of charging station utilization and user behavior. The approach presented here employs a novel co-evolutionary learning model for adaptive charging behavior. The simulation framework is tested and verified by means of a case study conducted in the city of Munich. The case study shows that the presented approach realistically reproduces charging behavior and spatio-temporal charger utilization.
... In other industrialized countries, the number of EVs has also risen considerably [2,3]. Electrification and on-demand services are two of the main driving forces within the current global automotive sector [4]. Increasing urbanization and stricter environmental regulations require a redesign of existing transportation systems [5]. ...
... Agent-based modeling is used when a system of vehicles is researched and the interaction between those vehicles is of interest [21]. Jäger et al. [4] realize an approach investigating the behavior of an electric taxi fleet with a combined approach using stochastic modeling and an agent-based simulation. The modeling results are subject to dynamic decision making. ...
There has been a significant increase in the sales of electric vehicles (EVs) in the United States and abroad in the last few years. Nevertheless, the overall adoption of these vehicles is hindered by range limits of EVs in conjunction with long charging times. In this context, it is essential to determine current energy demands and to predict future demand. This paper presents approaches for predicting energy consumption of EVs and discusses their eligibility for this purpose. Four modeling approaches (i.e., dynamic systems, neural networks, statistical models, physics-based models) have primarily been used in recent literature. In order to predict an EV's energy demand, several modeling techniques are combined to give an accurate prediction of the future energy consumption. For the battery EVs a combination of physics-based modeling and statistical modeling have shown to be an effective and efficient choice.
... Here, the authors dynamically adjust fleet sizes based on observed numbers of active taxis. Similarly, Jäger et al. (2017) simulate taxis that sample a desired shift duration at their respective operation start time. Long durations are possible to represent two-shift operations. ...
On-demand ride-pooling systems have gained increasing attention in science and practice in recent years. Simulation studies have shown an enormous potential to reduce fleet sizes and vehicle kilometers traveled if private car trips are replaced with ride-pooling services. However, existing simulation studies assume operation with autonomous vehicles, with no restrictions on operational tasks required when the vehicles are operated by manual drivers.
In this article, we simulate and evaluate the operational challenges of non-autonomous ride-pooling systems through driver shifts and breaks and compare their capacity and efficiency to autonomous on-demand services. Based on the existing ride-pooling service MOIA in Hamburg, Germany, we introduce shift and break schedules and implement a new hub return logic to perform the respective tasks at different types of vehicle hubs. This way, currently operating on-demand services are modeled more realistically and the efficiency gains of such services through autonomous vehicles are quantified.
The results suggest that operational challenges substantially limit the ride-pooling capacity in terms of served rides with a given number of vehicles. While results largely depend on the chosen shift plan, the presented operational factors should be considered for the assessment of current operational real-world services. The contribution of this study is threefold: From a technical perspective, it is shown that the explicit simulation of operational constraints of current services is crucial to assess ride-pooling services. From a policy perspective, the study shows the operational challenges of a ride-pooling service with non-autonomous vehicles and the potential of future autonomous services. Lastly, the paper adds to the literature a practical ride-pooling simulation use case based on observed real-world demand and shift data.
... This section describes the related work about the simulation modelling of the autonomous mobile systems and BSS. Jäger et al. (2017) illustrated the simulation testbed for autonomous mobility-on-demand systems using Anylogic simulation tool. This research mainly focused on autonomous mobility -on demand (AMoD) with autonomous shared taxis and provided necessary insights into modelling autonomous vehicles and travellers' interaction in the system. ...
In this modern era of transportation, bike-sharing systems dynamically increased their scope as one of the main mobility components in cities. Conventional bike-sharing systems significantly impact urban mobility, and the users have more interest in the utilization of bikes for short trips. However, the conventional bike-sharing systems have no dynamic rebalancing and less user friendly. To overcome these drawbacks, we proposed a next-generation autonomous bike-sharing system. This paper discusses the concept and system description for on-demand shared-use self-driving bikes service (OSABS). We need to investigate the economic viability and factors influencing the OSABS system. Furthermore, we present the simulation testbed with the technical implementation of the elements of the OSABS system. We demonstrate a typical agent-based simulation model in AnyLogic, representing self-driving cargo bikes, users, and the operational area of the OSABS system within the inner-city of Magdeburg.
... We distinguish between two approaches to fleet simulation. Object-oriented or agent-based models using a discrete-event framework were popularised in autonomous taxi fleet and general logistics simulation [39][40][41][42], but are also applied to bus fleets [14,15,43,44]. All objects in the simulation-vehicles, charging stations, depots, etc.-are simulated simultaneously in a shared environment with a central simulation clock, each object having its own individual state. ...
Bus operators around the world are facing the transformation of their fleets from fossil-fuelled to electric buses. Two technologies prevail: Depot charging and opportunity charging at terminal stops. Total cost of ownership (TCO) is an important metric for the decision between the two technologies; however, most TCO studies for electric bus systems rely on generalised route data and simplifying assumptions that may not reflect local conditions. In particular, the need to reschedule vehicle operations to satisfy electric buses’ range and charging time constraints is commonly disregarded. We present a simulation tool based on discrete-event simulation to determine the vehicle, charging infrastructure, energy and staff demand required to electrify real-world bus networks. These results are then passed to a TCO model. A greedy scheduling algorithm is developed to plan vehicle schedules suitable for electric buses. Scheduling and simulation are coupled with a genetic algorithm to determine cost-optimised charging locations for opportunity charging. A case study is carried out in which we analyse the electrification of a metropolitan bus network consisting of 39 lines with 4748 passenger trips per day. The results generally favour opportunity charging over depot charging in terms of TCO; however, under some circumstances, the technologies are on par. This emphasises the need for a detailed analysis of the local bus network in order to make an informed procurement decision.
... We distinguish between two approaches to fleet simulation. Object-oriented or agent-based models using a discrete-event framework were popularised in autonomous taxi fleet and general logistics simulation [39][40][41][42], but are also applied to bus fleets [14,15,43,44]. All objects in the simulationvehicles, charging stations, depots, etc. -are simulated simultaneously in a shared environment with a central simulation clock, each object having its own individual state. ...
Bus operators around the world are facing the transformation of their fleets from fossil-fuelled to electric buses. Two technologies prevail: Depot charging and opportunity charging at terminal stops. Total cost of ownership (TCO) is an important metric for the decision between the two technologies, however, most TCO studies for electric bus systems rely on generalised route data and simplifying assumptions that may not reflect local conditions. In particular, the need to re-schedule vehicle operations to satisfy electric buses’ range and charging time constraints is commonly disregarded. We present a simulation tool based on discrete-event simulation to determine the vehicle, charging infrastructure, energy and staff demand required to electrify real-world bus networks. These results are then passed to a TCO model. A greedy scheduling algorithm is developed to plan vehicle schedules suitable for electric buses. Scheduling and simulation are coupled with a genetic algorithm to determine cost-optimised charging locations for opportunity charging. A case study is carried out in which we analyse the electrification of a metropolitan bus network consisting of 39 lines with 4748 passenger trips per day. The results generally favour opportunity charging over depot charging in terms of TCO, however, under some circumstances, the technologies are on par. This emphasises the need for detailed analysis of the local bus network in order to make an informed procurement decision.
... For the simulation of the car-sharing system, an agent-based, discrete-event simulation approach was chosen. The simulation is written in JAVA and is derived from the fleet-simulation presented in [12]. It allows for the detailed simulation of multiple entities at reasonable runtimes. ...
Today, many cities provide a public transportation system capable of addressing the mobility needs of their citizens, reducing the dependency on private vehicles. Still, there are use cases where transport by private vehicle is more convenient, e.g., excursions to remote areas or the transportation of large goods. In recent years, car-sharing has evolved as an alternative to a privately owned vehicle, whereby people can rent a vehicle on the street for short durations compared to the traditional stationbased car rental options. BeeZero offered a car-sharing service in Munich (Germany) with a fleet of 50 hydrogen-powered vehicles. The business area was split up into several zones, distributed across the city area. The service focused on round-trip rentals, as each vehicle needed to be returned to the same zone where the rental began. This paper presents the development and implementation of an agent-based simulation model of the BeeZero car-sharing service. The model is validated with historical booking data and enables an analysis of the required walking distances to reach a vacant vehicle. By using data of a general travel survey, the potential and limits of the BeeZero carsharing service are evaluated using the simulation model for the city of Munich.
... A logging module stores simulation and debug information in a database. Further information about its architecture are published in [12]. The platform manages in total three agent types, namely office, taxi, and fleet management. ...
With recent policies such as the Clean Miles Standard in California and Lyft’s announcement to reach 100% electric vehicles (EVs) by 2030, the electrification of vehicles on ride-hailing platforms is inevitable. The impacts of this transition are not well-studied. This work attempts to examine the infrastructure deployment necessary to meet demand from electric vehicles being driven on Uber and Lyft platforms using empirical trip data from the two services. We develop the Widespread Infrastructure for Ride-hail EV Deployment model to examine a set of case studies for charger installation in San Diego, Los Angeles, and the San Francisco Bay Area. We also conduct a set of sensitivity scenarios to measure the tradeoff between explicit costs of infrastructure versus weighting factors for valuing the time for drivers to travel to a charger (from where they are providing rides) and valuing the rate of charging (to minimize the amount of time that drivers have to wait to charge their vehicle). There are several notable findings from our study: (1) DC fast charging infrastructure is the dominant charger type necessary to meet ride-hailing demand, (2) shifting to overnight charging behavior that places less emphasis on daytime public charging can significantly reduce costs, and (3) the necessary ratio of chargers is approximately 10 times higher for EVs in Uber and Lyft compared to chargers for the general EV owning public.
Shared mobility solutions improve traffic efficiency in cities. This paper addresses the design of an urban demand-responsive transit system operated by a fleet of autonomous modules as a public transportation service. We introduce a discrete event-based multi-agent simulation to reflect system dynamics at an operational level. Vehicles, customers, and intelligent stops are modeled as software agents. Each entity has its own decision-making and planning component. Vehicles compete for a limited number of order requests, and decentralized stops manage the matching between a customer and a prospective vehicle. Scheduling and routing events take place highly dynamically. A simulation study evaluates a demand-responsive operation as a replacement for a fixed-route bus service. Results show that a total demand of 2.3 million rides can be fulfilled with a fleet of around 43,000 six-seater modules. In comparison with the current bus system, the total energy consumption and the system’s utilization are improved, while transport times and distances result in values below the bus system’s performance and therefore must be optimized in further research.
—This paper presents a method for studying supply and demand in a taxi network in time and space by using the example of Munich. First, we introduce the necessary data collection that is linked to a fleet management system (FMS) operated by a local taxi agency and create a statistically sound database, which represents the mobility behavior on a trip level. Second, we derive key figures describing the city's taxi characteristics. Here both the temporal taxi supply and demand of 420 taxis over a period of 19 weeks is considered. As the taxi demand differs according to the city district, the investigated area has to be divided into various zones. We analyze their specific characteristics and describe key factors influencing taxi requests, such as weekdays, public holidays and number of point of interests within an area. Next, a model to predict a time-variant demand for individual districts is introduced. We classify the problem and choose an appropriate algorithm to forecast spatiotemporal taxi requests. As booking actions differ over time, a nonhomogeneous Poisson distribution is suitable for counting those events. In a final step, we validate the proposed model on a local and global scale. Our model helps to provide a better understanding of taxi fleet operations on a city scale. We use the suggested demand prediction as one input parameter for an agent-based fleet simulation that represents individual taxi movements. Index Terms—fleet analysis, fleet simulation, poisson model, taxi demand, intelligent vehicles, transportation I. MOTIVATION A process of radical change is taking place within the mobility sector. As we want to have a worthwhile future, we need to address various environmental and social challenges now [1]. Ongoing urbanization leads to a constant growth in the mobility demand in cities. Nevertheless, the existing traffic infrastructure has only limited resources. The increasing traffic cannot be absorbed to the same extent that an expansion of the infrastructure is possible, due to scarcity of space. This results in unavoidable traffic congestion with a rising number of traffic bottlenecks, especially in metropolitan areas.
The time has come: The Electromobility revolution has started.
How does this look? How fast will it take place?
Where will it start? Who is well-prepared for it? Who can be successful?
We consider a promising application for electric taxis (eTaxis) in Transportation Cyber-Physical Systems. The new rideshare scheme introduced herein, takes into consideration both the limited battery of eTaxis and the user requirements. In the proposed eTaxis sharing system, a passenger may share a taxi with others in order to enjoy a reduced fare, and can potentially transfer from one eTaxi to another before reaching her destination, as long as her total trip time is within the maximum she specifies to be tolerable. Transfers are restricted to only take place at the designated (safe and convenient) battery charging stations scattered around the city. When an eTaxi comes to a charging station to pick up/drop off a transfer passenger, the eTaxi's battery can be charged. In this paper, we address a new optimization problem called Transfer-Allowed Shared eTaxis (TASeT), whose goal is to schedule an electric taxi service and find optimal rideshare and transfer plans so as to maximize the system throughput in terms of the number of passengers served by the taxi service within a given time period. A Mixed Integer Programming formulation is presented to solve TASeT, along with an efficient greedy heuristic. Besides large-scale simulation, we also present a case study that utilizes real taxi traces collected from the city of Shanghai. Compared to the non-transferable taxi-sharing case, our solution could improve the number of served passengers and shared travel distance by 22% and 37% respectively during rush hours.
This study explores the challenge of the dynamic dispatching of taxis to the immediate passenger booking requests. In particular, the study leverages on a stable marriage assignment algorithm and applies it for dispatching taxis to passengers. The stable marriage algorithm was developed initially for matching men and women according to their preferences in polynomial time. The results of the custom built simulation model show that the taxi dispatching strategy based on the stable marriage matching improves the taxi operation performance in all observed indicators (taxi profit, number of served passengers, not-occupied and total taxi mileage and passenger waiting time) as compared to the standard first-come, first-served strategy.
Gives a unique up-to-date account of mainstream approaches to transport modelling with emphasis on the implementation of a continuous approach to transport planning. The authors discuss modern transport modelling techniques and their use in making reliable forecasts using various data sources. Importance is placed on practical applications, but theoretical aspects are also discussed and mathematical derivations outlined. -from Publisher
Electric vehicle (EV) taxis have been introduced into the public transportation systems to increase EV market penetration. Different from regular taxis that can refuel in minutes, EV taxis' recharging cycles can be as long as one hour. Due to the long cycle, the bad decision on the charging station, i.e., choosing one without empty charging piles, may lead to a long waiting time of more than an hour in the worst case. Therefore, choosing the right charging station is very important to reduce the overall waiting time. Considering that the waiting time can be a nonnegligible portion to the total work hours, the decision will naturally affect the revenue of individual EV taxis. The current practice of a taxi driver is to choose a station heuristically without a global knowledge. However, the heuristical choice can be a bad one that leads to more waiting time. Such cases can be easily observed in current collected taxi data in Shenzhen, China. Our analysis shows that there exists a large room for improvement in the extra waiting time as large as 30 min/driver. In this paper, we provide a real-time charging station recommendation system for EV taxis via large-scale GPS data mining. By combining each EV taxi's historical recharging events and real-time GPS trajectories, the current operational state of each taxi is predicted. Based on this information, for an EV taxi requesting a recommendation, we can recommend a charging station that leads to the minimal total time before its recharging starts. Extensive experiments verified that our predicted time is relatively accurate and can reduce the cost time of EV taxis by 50% in Shenzhen.
Learn how to employ JADE to build multi-agent systems! JADE (Java Agent DEvelopment framework) is a middleware for the development of applications, both in the mobile and fixed environment, based on the Peer-to-Peer intelligent autonomous agent approach. JADE enables developers to implement and deploy multi-agent systems, including agents running on wireless networks and limited-resource devices. Developing Multi-Agent Systems with JADE is a practical guide to using JADE. The text will give an introduction to agent technologies and the JADE Platform, before proceeding to give a comprehensive guide to programming with JADE. Basic features such as creating agents, agent tasks, agent communication, agent discovery and GUIs are covered, as well as more advanced features including ontologies and content languages, complex behaviours, interaction protocols, agent mobility, and the in-process interface. Issues such as JADE internals, running JADE agents on mobile devices, deploying a fault tolerant JADE platform, and main add-ons are also covered in depth. Developing Multi-Agent Systems with JADE: • Comprehensive guide to using JADE to build multi-agent systems and agent orientated programming. • Describes and explains ontologies and content language, interaction protocols and complex behaviour. • Includes material on persistence, security and a semantics framework. • Contains numerous examples, problems, and illustrations to enhance learning. • Presents a case study demonstrating the use of JADE in practice. • Offers an accompanying website with additional learning resources such as sample code, exercises and PPT-slides. This invaluable resource will provide multi-agent systems practitioners, programmers working in the software industry with an interest on multi-agent systems as well as final year undergraduate and postgraduate students in CS and advanced networking and telecoms courses with a comprehensive guide to using JADE to employ multi agent systems. With contributions from experts in JADE and multi agent technology.
The major obstacle to the wide acceptance of Electric Vehicles (EV) is the lack of a wide spread charging infrastructure. To solve this, the Chinese government has promoted EVs in public transportation. The operational patterns of EV taxis should be different from Internal Combustion Engine Vehicles (ICEV) taxis: EVs can only travel a limited distance due to the limited capacity of the batteries and an EV taxi may re-charge several times throughout a day. Understanding the status (e.g., operational patterns, driver income and charging behaviours) of EV taxis can provide invaluable information to policy makers. To our best knowledge, this is the first paper to understand EV taxis behavior patterns. We use real taxi GPS records data from a fleet with about 600 EV taxis operating in Shenzhen, China. We study the patterns from two aspects: operational behaviors and charging behaviors. The most important finding is: based on the net profits of both EV and ICEV taxis, which are derived from data, we find that commercial operation of an EV taxi fleet can be profitable in metropolitan area, when specific policies give advantages to EV taxis.
Electric vehicles (EVs) are gaining popularity amongst everyday commuters, but are still having a hard time penetrating the fleet vehicle market. In this paper, we investigate a model which does not require changes to current street infrastructure using charge sharing through inductive power transfer to determine if we can extend the driving range of EVs enough such that taxicab fleets can become EVs. Using fisheye state routing to coordinate rendezvous points, and a simple heuristic where cars exchange charge when there is excess available, we show that we can effectively increase the EV driving distance without needing to recharge. We use real data from the New York City Taxi and Limousine Commission to model taxicab usage behavior and the United States Department of Transportation Omnibus Survey to model commuter vehicle usage. Using this data, we demonstrate that an electric taxicab fleet could function with minimal disturbance (failure < 5%).
For the transportation sector, electromobility presents a chance both to abolish oil dependency and to open the possibility of using sustainable energy sources. In the case of taxis, the substitution of electric vehicles for conventional vehicles with an internal combustion engine may be especially favorable because the driving patterns involve several waiting periods, which can be used for recharging the battery. An infrastructure consisting of charging stations at taxi stands needs to be developed to ensure the energy supply of these taxis. Designing a charging infrastructure requires the development of an optimization method to maximize the economic benefit of the whole system, consisting of electric taxis (e-taxis) and charging stations (CSs). The number of charging stations should be kept as minimal as possible to reduce costs. Simultaneously, the energy supply for these taxis has to be ensured to enable high mileage and earnings. This study introduces an event-based simulation of the e-taxis' mileage and an economic analysis that evaluates the benefit of possible configurations of numbers of e-taxis and numbers of CSs per taxi stand. Because the number of possible configurations is high, an optimization algorithm is presented to reduce the calculation time and to find configurations with the highest economic efficiency. One result of the infrastructure optimization indicates that 445 CSs would be economically ideal for a taxi fleet in Munich, Germany, with 3,402 vehicles with a battery capacity of 20 kW-h.